Writers of Pro Football Prospectus 2008

20 Dec 2013

Let's Play Andrew Luck Whack-a-Mole

On Monday, I ran the list of leading defenders in QB hits for 2013 so far. Today, let's look at quarterbacks.

There are two ways for a quarterback to get knocked down: sacks, and QB hits after a pass is thrown. These are designated in the play-by-play in [brackets].

Nobody in the NFL gets knocked to the ground after a pass quite like Andrew Luck. Luck has 78 QB hits through 15 weeks, including plays cancelled by penalty. That's 25 more than any other quarterback in the league, a huge margin. This isn't a one-year thing, either. In his rookie year, Luck had 83 QB hits, which also led the league and was 22 more than any other quarterback. (Matt Ryan, incidentally, is second this year and was also second a year ago.)

However, Luck isn't the most frequently knocked down quarterback. First of all, he doesn't take as many sacks as some other quarterbacks; 31 sacks puts him in the middle of the pack. He also throws a lot of passes, so of course he's going to get knocked down a bit more.

If we look at knockdowns per pass, the leader is Geno Smith of the Jets, who has 47 hits and 44 sacks.

Two things to note about the following table:

1) As noted in the post about defenders and QB hits, official scorers are very inconsistent from stadium to stadium marking QB hits. Until we can clean that data up with game charting, some quarterbacks are artificially low because their home scorers almost never mark a QB hit, and others are probably artificially high.

2) I didn't have the time to go and do a proper rate stat that included all pass plays including those cancelled by penalty, so the "rate" listed here isn't actually the rate of knockdowns to pass plays. Nonetheless, it's close enough to be useful information.

QB Hits and Sacks, Weeks 1-15 2013
Player Team Hits Sacks Passes "Rate"
7-G.Smith NYJ 47 44 422 21.6%
7-C.Keenum HOU 39 19 269 21.6%
3-B.Weeden* CLE 36 27 293 21.5%
12-A.Luck IND 78 31 530 20.6%
3-R.Wilson SEA 43 37 394 20.3%
3-E.Manuel BUF 36 28 335 19.1%
10-R.Griffin WAS 45 39 495 17.0%
7-C.Ponder MIN 17 28 267 16.9%
3-C.Palmer* ARI 51 39 537 16.8%
6-J.Cutler CHI 38 13 308 16.6%
8-M.Schaub HOU 31 18 304 16.1%
7-C.Henne JAC 38 33 451 15.7%
Player Team Hits Sacks Passes "Rate"
17-R.Tannehill MIA 36 53 570 15.6%
2-T.Pryor** OAK 11 29 262 15.3%
9-N.Foles** PHI 21 22 289 14.9%
10-K.Clemens STL 12 19 209 14.8%
2-M.Ryan ATL 53 35 597 14.7%
17-J.Campbell* CLE 24 12 245 14.7%
8-M.Glennon TB 24 32 382 14.7%
5-J.Flacco BAL 38 45 567 14.6%
4-R.Fitzpatrick TEN 27 19 315 14.6%
10-E.Manning NYG 39 36 521 14.4%
12-J.McCown CHI 20 12 232 13.8%
12-T.Brady NE 45 39 611 13.7%
Player Team Hits Sacks Passes "Rate"
8-S.Bradford STL 23 15 277 13.7%
7-C.Kaepernick SF 19 34 396 13.4%
12-A.Rodgers GB 17 18 268 13.1%
11-A.Smith KC 31 34 513 12.7%
9-D.Brees NO 46 30 604 12.6%
1-C.Newton** CAR 17 40 460 12.4%
7-B.Roethlisberger** PIT 22 40 560 11.1%
9-M.Stafford DET 43 16 576 10.2%
17-P.Rivers SD 24 26 508 9.8%
9-T.Romo DAL 20 33 543 9.8%
14-A.Dalton CIN 20 28 539 8.9%
16-M.Cassel MIN 8 9 204 8.3%
18-P.Manning DEN 27 16 598 7.2%
* QB hits total may be artificially high due to overzealous home scorers.
** QB hits total may be artificially low due to inconsistent home scorers.
Note: This is not the exact rate of knockdowns to pass plays because QB Hits and Sacks totals
include plays cancelled by penalty, but Passes total does not.

Posted by: Aaron Schatz on 20 Dec 2013

37 comments, Last at 21 Jan 2014, 8:21pm by Fake Pandora Bracelets

Comments

1
by tuluse :: Fri, 12/20/2013 - 1:47pm

Thanks Aaron. Some really interesting data here, but I'm not sure how to interpret it. One random thing I noticed, Cam Newton is sacked about twice as often as Jay Culter but takes a hit at about 1/3 as often.

I wonder if the difference between sack and hit rates would tell us anything.

3
by Nathan :: Fri, 12/20/2013 - 2:34pm

I'd imagine with Newton a lot of those sacks are actually carries for negative yards and scrambles. Stands to reason an athletic QB would take off as the pocket collapses and look to make a play with his legs rather than stand in, throw a pass and take a hit. Would explain the disparity.

4
by RickD :: Fri, 12/20/2013 - 2:40pm

I think it would tell you that Newton is more likely to get the pass away before getting hit.

5
by tuluse :: Fri, 12/20/2013 - 2:55pm

This includes hits after the pass is thrown.

7
by Eddo :: Fri, 12/20/2013 - 3:41pm

Wouldn't it be the other way around?

If Newton is hit less, but sacked more, it means his sacks per total contact is higher.

10
by Nathan :: Fri, 12/20/2013 - 3:50pm

When someone is sacked does it count as a sack + a hit or just as a sack?

12
by CBPodge :: Fri, 12/20/2013 - 4:04pm

Just a sack, otherwise it would be impossible to have fewer hits than sacks.

15
by Nathan :: Fri, 12/20/2013 - 4:22pm

What if you break a tackle? Does that count as a hit? Cause if not, Eddo's "sacks per total contact" point is incorrect.

17
by tuluse :: Fri, 12/20/2013 - 4:40pm

A "hit" means the qb was knocked to the ground. If a qb is knocked to the ground with the ball, that's a sack.

8
by CBPodge :: Fri, 12/20/2013 - 3:48pm

So I decided to put this data into a spreadsheet and have a quick fiddle. The first thing I wanted to look at was ratio hits to sacks. The players at either end seem to (unsurprisingly) fit in with their anecdotal "style". The bottom 5 guys in hit to sack ration (so the guys who take most sacks as percentage of total hits) are:

Pryor, Newton, Big Ben, Kaepernick and Romo.

All of those guys are seen as the sort of guys who are thought to likely run around and try to make things happen when under pressure.

Top top 5 guys are Cutler, Stafford, Luck, Keenum and Campbell. If you take out the two part time starters (Keenum and Campbell) you get Manning and Brees as the next two. Those all feel like guys who are known to sit in the pocket and trust their arm.

Basically, the guys who you think run around a lot to avoid sacks get sacked more regularly than they get hit, the guys who don't avoid the rush but try to get the ball out instead get hit more than they get sacked. That feels like one of those things that confirms exactly what you expect, but is sort of interesting simply because of that.

The overall hit-to-sack ratio for the league is 1.11. That's a little surprising - the numbers of hits that are sacks and hits that aren't sacks are closer than I'd have expected.

The average hit rate is 7.5%. The average sack rate is 6.7%. One interesting here: Matt Stafford's sack rate is 2.8%, second lowest in the league. But his hit rate is 7.5%, exactly league average. So I guess you could argue that this shows that Detroit's low sack rate is mainly a function of Stafford being excellent at getting rid of the ball at the last moment, rather than the line holding people off him.

14
by tuluse :: Fri, 12/20/2013 - 4:09pm

Is this essentially measuring how quick the release is (plus how quick the decision making is [but that was already partially measured with just sacks])?

16
by Perfundle :: Fri, 12/20/2013 - 4:36pm

"Matt Stafford's sack rate is 2.8%, second lowest in the league. But his hit rate is 7.5%, exactly league average."

But it's not hit rate that you should be looking at; it's hit + sack rate, which is what's listed, and in that Stafford's line looks very good. Given that fact, you can flip around your reasoning and say that Detroit's low sack rate is mainly a function of Stafford not letting routes develop downfield and throwing passes too quickly and with poor form when he feels the slightest pressure, as in his last interception on Monday.

18
by JoeyHarringtonsPiano :: Fri, 12/20/2013 - 4:47pm

This.

The line has been surpisingly good this year. Stafford has always had a quick release, and gets rid of the ball quickly, which is usually good. But there have been times where it would have been better for him to take a sack rather than try to force a play that wasn't there, by throwing while falling backwards or some other nonsense. Being tied with Manning for fewest sacks is somewhat deceiving.

24
by CBPodge :: Fri, 12/20/2013 - 5:29pm

My point was that if you just looked at Stafford's sack rate you'd say "wow, he and the line are doing a fantastic job at avoiding pressure on him", but then the additional data suggests that ain't quite the case.

To your point, I dunno, because "slightest pressure" wouldn't likely lead to a hit (or if it did, it'd then lead to a roughing the passer penalty).

I guess the thing to sort of close the loop here might be seeing how players' DVOA when being hit is compared to DVOA when not hit? I recall from the last book that every QB's DVOA was significantly lower when pressured, but it would be interesting to see whether there are certain passers who are making the decisions to let the ball go when hit but still being (relatively) effective?

In reality, I suspect the sample size is too small. I think this stat might be similar to one of the interpretations of stuff like stop rate, in that it can be seen as an indication of how a player plays, rather than his actual effectiveness in that style?

29
by Perfundle :: Fri, 12/20/2013 - 6:25pm

Well, if you want a bigger sample size, look at the last three years. Detroit fans have said that the OL was at its best in 2011, but Stafford was sacked significantly more in 2011 compared to 2012. However, he did the best in the year he was sacked the most, which suggests that when there was little pressure on him, he could afford to survey the field and pass with good form, whereas he tends to go to pieces in his mechanics when he feels pressure and rushes throws to avoid sacks.

31
by JoeyHarringtonsPiano :: Fri, 12/20/2013 - 10:53pm

This website usually charts DVOA when pressured and not pressured. Staffords DVOA when pressured in 2011 and 2012 was among the league leaders (like +30 DVOA points higher than league average). However his DVOA when not pressured is closer to league average. His wierd mechanics sometimes allows him to make impossible throws when a guys in his face. However, sometimes the wierd mechanics rear their ugly head when he has a clean pocket, making his unpressured performance not reach its full potential.

32
by Hurt Bones :: Fri, 12/20/2013 - 11:22pm

I think that's weird.

33
by JoeyHarringtonsPiano :: Sat, 12/21/2013 - 1:09pm

Seems like it on the surface, but remember everyone's DVOA looks terrible if you only look at the plays when under pressure (Pretty much everyone gets negative numbers). Guys like Manning, Brady, and strangely, Cutler and Stafford are less negative than everyone else. The difference is when Manning and Brady are unpressured, their DVOA is astronomically high, while Stafford's is only "sorta" high. This suggests he doesn't get bothered that much by pass pressure, but when he's not pressured, he doesn't take advantage of it like he should.

34
by Hurt Bones :: Sat, 12/21/2013 - 1:24pm

A little unusual, but I was thinking more in terms of English spelling is weird.

Sheila, the counterfeiter, was seized by a weird thought at the weir.

19
by mehllageman56 (not verified) :: Fri, 12/20/2013 - 4:49pm

The top 7 guys are in their first two years in the league. Perhaps inexperience in reading defenses leads to sacks. Of course, I doubt Wilson really has a problem reading defenses; his line has been injured all year.

20
by tuluse :: Fri, 12/20/2013 - 4:58pm

On the other hand the 5 guys listed next are all long time vets.

2
by PAX (not verified) :: Fri, 12/20/2013 - 2:08pm

Impressive that Keenum accomplished so much abuse in so few games.

6
by Ben :: Fri, 12/20/2013 - 3:01pm

The other thing with Luck is that he pulls off a Houdini escape what seems like at least once a game. The Colts' O-Line is truly horrific.

9
by CBPodge :: Fri, 12/20/2013 - 3:50pm

From the fiddling with numbers I did (above), Luck's hit rate is 14.7% (highest in the league), but his sack rate is 5.8% (below the average of 6.7%). That doesn't feel like a great disparity (from the line's perspective).

30
by Ben :: Fri, 12/20/2013 - 7:11pm

I haven't counted these via game tape or anything, but I'd guess there are 9 or 10 times this season when Luck has managed to get out of what looked to be a sure sack, through no help of his o-line. Adding those in would certainly raise that sack rate to well above average, and, in my mind, validate the argument that the O-Line is terrible.

11
by Jon Goldman (not verified) :: Fri, 12/20/2013 - 3:50pm

The Bears' QBs are not first on this list.

I know this is sort of out there, but that sentence feels wonderful to say.

22
by whckandrw (not verified) :: Fri, 12/20/2013 - 5:02pm

I know how you feel. Also the gap between Cutler and McCown tells us a lot. It feels like Cutler is getting the ball out faster this year but he still holds it longer than he should.

27
by tuluse :: Fri, 12/20/2013 - 5:53pm

He has a lower sack rate than McCown. He might be holding the ball longer, but it's leading to hits, not sacks.

13
by Riceloft :: Fri, 12/20/2013 - 4:04pm

What does the data for Brian Hoyer look like? I know its a small sample size, but I'd be very interested!

21
by MJK :: Fri, 12/20/2013 - 4:59pm

Think about what a knockdown rate of, say, 20% means. It means one pas play in five one or more 250lb + men running full tilt are going to knock you on your rear. How many passes per game does a QB attempt, on average? Ouch!

Of course, other positions (like RB) have a nearly 100% knockdown rate. No wonder there are concussion issues!

23
by whckandrw (not verified) :: Fri, 12/20/2013 - 5:05pm

The big takeaway of this list is "quarterbacks have a larger influence on quarterback hits than offensive lines."

25
by CBPodge :: Fri, 12/20/2013 - 5:31pm

My name is Andrew Luck and I disagree with your conclusion.

But yeah, it does seem largely another data point.

I think it would be interesting to tie it in with pressures as well, to see how different QBs handle pressure - who gets rid of the ball before a pressure turns into a hit or sack, who takes the hit, who takes the sack?

26
by intel_chris (not verified) :: Fri, 12/20/2013 - 5:49pm

Along with the hit & sack rates, it would be good to see the numbers of hurries. Even more telling might be incompletions and interceptions on hurries and hits. For example, in the Stafford case, if he really is just hurrying, but making poor throws and choices, that might show up in the number of incompletions+interceptions/hurries+hits. Of course, the more you dice the statistics, the more likely you are going to end up with too small of samples.

28
by Aaron Schatz :: Fri, 12/20/2013 - 6:20pm

Again, remember my note about the official scorers. One reason Newton and Roethlisberger are low is that the official scorers in CAR and PIT mark significantly fewer QB hits than average.

35
by cheap Detroit Lions Jerseys (not verified) :: Tue, 01/21/2014 - 8:21pm

Simply want to say the article is usually as astounding. The clearness on your own put up is just cool and i could think you

35
by Pandora Outlet Store (not verified) :: Tue, 01/21/2014 - 8:21pm

Hey! Do you have Twitter? I

35
by Fake Pandora Bracelets (not verified) :: Tue, 01/21/2014 - 8:21pm

Its for example you discover my thoughts! You may actually grasp a lot approximately this particular, like a person wrote the actual guide inside it or some thing. I feel that you just could perform with a few percent in order to force the actual message house a bit, however besides that, this is excellent blog. An excellent read. I